Fri, 12/01/2023 - 00:37
Sometimes I think I see indications of light pollution in spectra of faint stars I take with longer exposures using a slitless approach. I know I have some high pressure sodium vapor lights a mile away in the valley below me. I don't recall seeing the spectrum of a light polluted sky taken and presented recently. I know that if you take spectra of a light pollution source you could get emission lines or a diffuse spectrum depending on what it is... but summing up all sources and taking a long exposure spectrum of the night sky is another matter and I'm wondering what it would look like? Would emission bands in light sources show up in a sky spectrum as absorption features?
First post by newbie member,
Rick
Hi Rick,
You do not get a spectrum of the night sky with slitless systems, just a smooth background. Light pollution certainly limits how deep you can go with slitless systems. This is one of the advantages of slit spectrographs where the slit keeps out most of the sky background. The remaining sky background then forms a spectrum with broad bands and narrow emission lines depending on the light pollution (both artificial and natural). You can see an example of my sky spectrum here in this deep low resolution spectrum image of a magnitude 18 supernova taken with a slit spectrograph (A modified ALPY)
Note the image are negative so emission is black
https://britastro.org/observations/observation.php?id=20231119_171541_a5ef1da74745623d
Cheers
Robin
My rural night sky is quite dark but It is possible to do spectroscopy of faint objects with a slit spectrograph even with high levels of light pollution. See these examples by Christian Buil
http://www.astrosurf.com/buil/pollution/measure.htm
My rural night sky is quite dark but It is possible to do spectroscopy of faint objects with a slit spectrograph even with high levels of light pollution. See these examples by Christian Buil
http://www.astrosurf.com/buil/pollution/measure.htm
Robin,
Although I am currently using a grating/prism rig with about R=250 on my small refractor, I sometimes see a broad and deep absorption featureor features running approximately between 5800 and 6100 angstroms. Otherwise, the spectrum shape and features usually superimposes quite well on the reference spectrum for the star I'm imaging. This happens most often when I'm imaging a middling magnitude star to the west and somewhat low in the sky. I have thought that this was somehow caused by light pollution as there are some high pressure sodium lights in the general area. I have taken the trouble to put a DIY slit on my rig and confirmed the high pressure sodium emission lines in these distant but bright light sources in the valley below me.
Is it likely that the abnormal absorption troughs that are ruining these spectra are unrelated to light pollution but have another cause? The issue tends to occur when I'm facing west and lower in the sky than usual, and when exposure times are longer than usual and the spectra are a bit noisier than I would like.
Thanks for the references, I will read them with interest.
Rick
In slitless spectra the sky background is a combination of zero orders and multi-order spectra from every point in the sky, even those outside the camera field diffracted by the grating into the field of view. You will get gradients in the sky background as any normal imaging and possibly you are seeing broad bands from diffuse light pollution sources outside the field, diffracted into the field. With slitless systems though you cannot relate them to specific wavelengths as they do not come from the same point in the sky as the star.
They will be removed in processing however by the sky background subtraction and response correction so (except for the degradation in signal/noise) they don't affect the accuracy of the spectrum. This example shows the sky background of a wide slit system (effectively almost slitless) and what happens to it as you close the slit.
http://www.threehillsobservatory.co.uk/astro/spectroscopy_4.htm
You said " I sometimes see a broad and deep absorption feature or features running approximately between 5800 and 6100 angstroms"
Can you clarify please? Do you actually see these in the sky background or do you only see them in the final spectrum ? There are many reasons for broad bands appearing in the final spectrum which are not related to light pollution. With slitless systems for example where flats cannot be used you see ripples in the response of the camera sensor. These will appear in both the reference and target spectra however so should be included in the response There will also be telluric bands from our own atmosphere which are most intense at wavelengths around H alpha and higher but there are also some around the Na D line and a broad band due to O3 which will be more obvious at high air mass. You can see them here for example in Christian Buil's measurement of the atmospheric transmission here (near the bottom of the page)
http://www.astrosurf.com/buil/atmosphere/transmission.htm
You can remove the Telluric bands if you like either by including them in the response curve or by subtracting them using a scaled template but most of the time they are just left in (AAVSO request they are left in for their database)
You say "The issue tends to occur when I'm facing west and lower in the sky than usual,"
This would be the case if they are telluric absorption bands. (Note if you are using a reference star for response correction (flux calibration) It is more important at high air mass that your reference star is at a closely similar air mass or any difference in air mass is corrected for
So, I had a calibration error which I believe was caused by the sensor/grating distance being slightly different than when my instrument response curve was done (i.e. 3.2 angstrom/pixel dispersion vs 2.9 angstrom/pixel dispersion). This would have been caused by turning the grating/prism assembly to get a horizontal spectrum.
When I did a calibration using the telluric band at 6869 angstroms (obvious in my spectra) and my zero order image and got the 2.9 angstrom/pixel dispersion everything started to make sense. I am confident now that the "strange band" I thought was light pollution effect was simply the Mg triplet Fe I band. What I had taken for a H Balmer band turned out to really be Calcium. This star had minimal Balmer bands and was K type.
I am reasonably sure that is what happened.
Rick